Intelligent diagnosis apparatus and method for partial variable value monitoring in power distribution station room

Abstract

An intelligent diagnosis apparatus and method for partial variable value monitoring in a power distribution station room. Each subspace is obtained for a received partial discharge value, a trend change in the partial discharge value of each subspace is parsed, the distortion amplitude of the partial discharge value at each sampling time point is established, and on the basis of the distortion amplitude, the distortion stability of the partial discharge value at each sampling time point in each subspace is obtained, wherein the distortion stability reflects the stability of the distortion amplitude of the partial discharge value at each sampling time point along with a change in the capacity of the subspace, and the trend change situation of serval subspaces is involved; on the basis of the distortion stability, the distortion accuracy of the partial discharge value at each sampling time point in each subspace is obtained, so as to obtain a correct distortion amplitude of the partial discharge value at each sampling time point; and the correct distortion amplitude is used to remove a distorted value, such that the partial discharge value with noise removed is compared with a defined threshold value. The present invention has the effects of high accuracy of partial discharge values, high precision in comparing the partial discharge values with a defined threshold value, and a high level of safety.

Description

Intelligent diagnosis device and method for monitoring partial discharge value of power distribution station room TECHNICAL FIELD

[0001] The application belongs to the technical field of monitoring partial discharge value of power distribution station room, and particularly relates to an intelligent diagnosis device and method for monitoring partial discharge value of power distribution station room. BACKGROUND

[0002] The power distribution station room is a power distribution station building, which includes a switch station, a power distribution room, a ring network room, a ring network box and a box-type substation. According to statistics of the power distribution station room, partial discharge, that is, partial discharge, is an important reason for the final insulation breakdown of high-voltage electrical equipment of the power distribution station room and an important indicator of insulation deterioration. In order to ensure safety, partial discharge monitoring of the power distribution station room needs to be performed.

[0003] To perform partial discharge monitoring on the power distribution station room, the prior art with the patent publication number "CN212258243US001" is often used to perform partial discharge monitoring on the gas insulated high-voltage cabinet of the switch station of the power distribution station room. Specifically, the partial discharge monitoring terminal is connected to the monitoring server via a communication device. The partial discharge value of the gas insulated high-voltage cabinet obtained by sampling is transmitted to the monitoring server via the communication device. The monitoring server compares the collected partial discharge value with the defined threshold value. If the collected partial discharge value is higher than the defined threshold value, the monitoring server will alarm the alarm connected thereto.

[0004] However, because the partial discharge value is affected by internal and external noise elements during transmission to the monitoring server via the communication device, the partial discharge value often has noise value-induced disturbance. Therefore, before comparing the partial discharge value with the defined threshold value, the partial discharge value needs to be calibrated to remove noise to ensure the accuracy of the collected partial discharge value.

[0005] In view of the partial discharge values sampled in chronological order, the partial discharge values often show a certain trend in chronological order. Therefore, the distortion of the partial discharge value can be reflected by the difference between the individual partial discharge value and the trend of the time point. Here, the partial discharge value in the constant-capacity sub-area is often used as a unit of measurement in view of the trend of the time point. However, because the trend of the partial discharge value of different numbers has significant differences, the partial discharge value in the constant-capacity sub-area often causes numerical loss, which weakens the accuracy of the individual partial discharge value and is not conducive to comparing the partial discharge value with the defined threshold value. SUMMARY

[0006] In order to solve the defects in the prior art, the present application provides an intelligent diagnosis device and method for monitoring local variation values in a power distribution station room. The local discharge values of each sub-space are obtained by collecting the local discharge values. The trend variation of the local discharge values of each sub-space is analyzed. The distortion amplitude of the local discharge values at each sampling time point is established. The difference between the local discharge values at each sampling time point and the overall trend variation in the sub-space is involved. The amplitude of the local discharge values at each sampling time point disturbed by the noise values is reflected. The distortion stability of the local discharge values at each sampling time point in each sub-space is obtained according to the distortion amplitude. The distortion stability of the local discharge values at each sampling time point in each sub-space is reflected with the change of the capacity of the sub-space. The trend variation condition of several sub-spaces is involved. The distortion correctness of the local discharge values at each sampling time point in each sub-space is obtained according to the distortion stability. The distortion stability of the local discharge values in each sub-space and the regression offset of the local discharge values in each sub-space are involved. The safety of the local discharge value distortion amplitude is improved. The correct distortion amplitude of the local discharge values at each sampling time point is obtained. The distortion values are removed by using the correct distortion amplitude. The local discharge values after removing the noise are compared with the defined critical value. If the local discharge values after removing the noise are higher than the defined critical value, the alarm connected to the monitoring server alarms. The present application has the advantages of good local discharge value correctness, good precision of the local discharge values compared with the defined critical value, and good safety.

[0007] The present application uses the following technical solutions.

[0008] An intelligent diagnosis method for monitoring local variation values in a power distribution station room, comprising:

[0009] The local discharge monitoring terminal transmits the local discharge values of the components in the power distribution station room obtained by sampling to the monitoring server through the communication equipment. The monitoring server removes the noise from the collected local discharge values. Then, the local discharge values after removing the noise are compared with the defined critical value. If the local discharge values after removing the noise are higher than the defined critical value, the monitoring server alarms the alarm connected thereto.

[0010] The method for removing the noise from the collected local discharge values by the monitoring server, comprising:

[0011] Step 1: Collect the local discharge value queue and perform the pre-treatment.

[0012] Step 2: Obtain the distortion amplitude of the local discharge values at each sampling time point in each sub-space after the pre-treatment. Obtain the distortion stability of the local discharge values at each sampling time point in each sub-space according to the distortion amplitude. Obtain the distortion correctness of the local discharge values at each sampling time point in each sub-space according to the distortion stability. Obtain the correct distortion amplitude of the local discharge values at each sampling time point.

[0013] Step 3: Obtain the correct distortion amplitude of the local discharge values at all sampling time points, and thus select all the distortion values.

[0014] Step 4: Based on the partial discharge value after removing the distortion value, the final goal is to compare the partial discharge value after clutter removal with the defined critical value.

[0015] Preferably, in Step 1, the partial discharge monitoring terminal samples the partial discharge values ​​of the components in the substation room and transmits the partial discharge values ​​and their sampling times to the monitoring server in real time according to the order of sampling times. The monitoring server arranges the received partial discharge values ​​into a partial discharge value queue according to the order of their sampling times.

[0016] Preferably, in Step 1, given all the collected partial discharge values, a moving average method is initially used to perform preprocessing on the partial discharge values ​​to obtain the partial discharge value at each sampling point after preprocessing.

[0017] Preferably, Step 2 specifically includes:

[0018] For each partial discharge value after preprocessing in the partial discharge value queue, several adjacent partial discharge values ​​are selected to form a partial space. Then, regression is performed on all partial discharge values ​​in the partial space, using the sampling time point as the X coordinate and the partial discharge value as the Y coordinate. The RANSAC method is used to obtain the regression line of all partial discharge values ​​in the partial space. Next, the median in the partial space is removed, and the RANSAC method is used to obtain the regression line of the remaining partial discharge values. The regression offset of a pair of regression lines is calculated, that is, the regression offset of the regression line of the partial discharge values ​​in the partial space and the regression offset of the regression line of the remaining partial discharge values, which are defined as regression offset one and regression offset two, respectively.

[0019] Preferably, the regression offset is the amount obtained by subtracting the partial discharge value on the regression line at the sampling time point from the partial discharge value obtained by sampling at the sampling time point;

[0020] Corresponding to the partial discharge value at each sampling point after preprocessing in the partial discharge value queue, the partial discharge values ​​are sequentially taken from the left and right sides of the partial discharge value queue. Each partial discharge value is used as the partial discharge value distribution space at each sampling point, and is defined as the highest partial space.

[0021] Preferably, Step 2 further includes: establishing the distortion amplitude of the partial discharge value at each sampling time point based on the regression offset, the equation of which is: Within the equation, Representing the The distortion amplitude of the partial discharge value at the sampling point in the highest spatial distribution. Representing the The regression offset of the partial discharge value at the sampling time point in the highest distribution space is one. Representing the The regression offset of the partial value at the sampling point in the highest distribution space is two. representing the number of local emission values contained in the highest sub-space, representing the use of the Z-score method, performing standardization.

[0022] Preferably, Step 2 further comprises: defining sub-spaces of different capacities for each local emission value of the sampling time point, the number of local emission values contained in the sub-spaces being , where represents the query frequency, represents the query direction, the value of is two, the value of is two, three, four, five, six, seven, eight, nine, ten, eleven, ten sub-spaces of different capacities are identified for each local emission value of the sampling time point through ten queries;

[0023] The distortion amplitude of each local emission value of the sampling time point in each sub-space is obtained by using the same operation method as the distortion amplitude of the local emission value in the highest sub-space.

[0024] Preferably, Step 2 further comprises: for the local emission value of the first sampling time point, establishing the distortion smoothness in each sub-space, the equation being: in the equation, represents the distortion smoothness of the local emission value of the first sampling time point in the first sub-space, represents the distortion amplitude of the local emission value of the first sampling time point in the first sub-space, represents the distortion amplitude of the local emission value of the first sampling time point in the first sub-space, represents the preset operation span.

[0025] Preferably, the distortion correctness of each local emission value of the sampling time point in each sub-space is established, the equation being: in the equation, represents the distortion correctness of the local emission value of the first sampling time point in the first sub-space, represents the distortion smoothness of the local emission value of the first sampling time point in the first sub-space, represents the first sampling time point of the local emission value, The regression offset of all partial discharge values ​​within each subspace. Representing the The partial discharge value at the sampling time point The regression offset of all partial discharge values ​​in each subspace. Representing the The partial discharge value at the sampling time point The number of partial values ​​contained in each subspace. Representing the The partial discharge value at the sampling time point The number of partial values ​​contained in each subspace. The representative used the Z-score method to... Implement standardization.

[0026] Preferably, the correct distortion amplitude of the partial discharge value at each sampling point is established by combining the distortion amplitude and distortion accuracy, and the equation is: Within the equation, Representing the The correct distortion amplitude of the partial discharge value at the sampling point. Representing the The partial discharge value at the sampling time point is at the first The distortion accuracy of each subspace, Representing the The partial discharge value at the sampling time point is at the first The distortion amplitude of each spatial component Representing the The number of partial discharge values ​​in the sampling time space.

[0027] Preferably, Step 3 specifically includes: first, applying the Z-score method to standardize the correct distortion amplitude of the partial discharge values ​​at all sampling points to obtain a standardized value, and then raising the standardized value above a certain limit. The partial discharge value is identified as a distortion value.

[0028] Preferably, Step 4 specifically includes: facing all partial discharge values ​​after preprocessing, clearing all distortion values, then using the EM algorithm to fill in the partial discharge values ​​at the sampling time points where the partial discharge values ​​have been cleared, then using the filled partial discharge values ​​as the partial discharge values ​​after clutter removal, and comparing the partial discharge values ​​after clutter removal with a defined threshold value. If the partial discharge values ​​after clutter removal are higher than the defined threshold value, the monitoring server will trigger the alarm connected to it.

[0029] A smart diagnostic device for monitoring local variable values ​​in a substation room, comprising:

[0030] The partial discharge monitoring terminal arranged on the component of the power distribution station room is connected with the monitoring server through the communication device, the monitoring server is connected with the server, the partial discharge monitoring terminal is used for transmitting the partial discharge value of the component of the power distribution station room sampled to the monitoring server through the communication device, the monitoring server compares the partial discharge value collected with the defined critical value after executing the wave removal, if the partial discharge value after the wave removal is higher than the defined critical value, the monitoring server alarms the alarm connected therewith;

[0031] The module running on the monitoring server comprises:

[0032] The preceding processing module is used for collecting the partial discharge value queue and executing the preceding processing;

[0033] The smoothness module is used for obtaining the distortion amplitude of the partial discharge value of each sampling time point in each partial space after the preceding processing, obtaining the distortion smoothness of the partial discharge value of each sampling time point in each partial space according to the distortion amplitude, obtaining the distortion correctness of the partial discharge value of each sampling time point in each partial space according to the distortion smoothness, and obtaining the correct distortion amplitude of the partial discharge value of each sampling time point;

[0034] The selection module is used for obtaining the correct distortion amplitude of the partial discharge value of all sampling time points, thereby selecting all the distortion values;

[0035] The clearing module is used for finally comparing the partial discharge value after the wave removal with the defined critical value according to the partial discharge value after the distortion values are removed.

[0036] The beneficial effects of the present application are that, compared with the prior art, the technical effects of the present application include:

[0037] The trend change of the partial discharge value of each sub-space is analyzed, the distortion amplitude of the partial discharge value of each sampling time point is established, the difference between the partial discharge value of each sampling time point and the overall trend change in the sub-space is related, the amplitude of the disturbance of the partial discharge value of each sampling time point by the noise value is reflected, the distortion stability of the partial discharge value of each sampling time point in each sub-space is obtained according to the distortion amplitude, the stability of the distortion amplitude of the partial discharge value of each sampling time point with the change of the capacity of the sub-space is reflected, the trend change condition of a plurality of sub-spaces is related, the distortion correctness of the partial discharge value of each sampling time point in each sub-space is obtained according to the distortion stability, the distortion stability of the partial discharge value in each sub-space and the regression offset of the partial discharge value in each sub-space are related, the safety of the distortion amplitude of the partial discharge value is improved, and the correct distortion amplitude of the partial discharge value of each sampling time point is obtained, the distortion value is removed by using the correct distortion amplitude, and the partial discharge value after removing the noise is compared with the defined critical value. If the partial discharge value after removing the noise is higher than the defined critical value, the monitoring server alarms the alarm connected thereto. The application has the effects of good partial discharge value correctness, good precision of the partial discharge value compared with the defined critical value, and good safety. BRIEF DESCRIPTION OF DRAWINGS

[0038] Fig. 1 is a flow chart of the partial variation value monitoring intelligent diagnosis method of the power distribution station room described in the application;

[0039] Fig. 2 is a partial structure schematic diagram of the partial variation value monitoring intelligent diagnosis device of the power distribution station room described in the application. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical scheme and advantages of the application more clear, the technical scheme of the application will be clearly and completely expressed below by combining the drawings in the embodiments of the application. The embodiments expressed in the application are only a part of the embodiments of the application, not all the embodiments. According to the spirit of the application, other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the application.

[0041] As shown in Fig. 1, the partial variation value monitoring intelligent diagnosis method of the power distribution station room described in the application comprises:

[0042] The partial discharge value of the component of the power distribution station room sampled by the partial discharge monitoring terminal is transmitted to the monitoring server through the communication equipment, the monitoring server removes the noise of the received partial discharge value, and then compares the partial discharge value after removing the noise with the defined critical value. If the partial discharge value after removing the noise is higher than the defined critical value, the monitoring server alarms the alarm connected thereto;

[0043] The method for removing the noise of the received partial discharge value by the monitoring server comprises:

[0044] Step1: Collecting the partial discharge value queue and performing pre-processing;

[0045] In the preferred but non-limiting embodiment of the present application, in Step1, the partial discharge monitoring terminal is used to sample the partial discharge value of the components in the substation room, and the partial discharge value is transmitted to the monitoring server in real time according to the sampling time point in the order of the sampling time point. The monitoring server arranges the collected partial discharge values in the order of the sampling time point to form a partial discharge value queue, thereby achieving the purpose of collecting the partial discharge value queue.

[0046] In the preferred but non-limiting embodiment of the present application, in Step1, the sampling period in the present application is 2s, that is, the partial discharge monitoring terminal samples the partial discharge value of the components in the substation room every 2s. For all the collected partial discharge values, that is, all the partial discharge values in the partial discharge value queue, the sliding average method is used to perform pre-processing on the partial discharge values, that is, to perform noise removal processing in advance, and to obtain the partial discharge value of each sampling time point after pre-processing.

[0047] Step2: Obtaining the distortion amplitude of the partial discharge value of each sampling time point in each sub-space, obtaining the distortion stability of the partial discharge value of each sampling time point in each sub-space according to the distortion amplitude, obtaining the distortion correctness of the partial discharge value of each sampling time point in each sub-space according to the distortion stability, and obtaining the correct distortion amplitude of the partial discharge value of each sampling time point;

[0048] In the preferred but non-limiting embodiment of the present application, Step2 specifically includes: the present application starts to collect the partial discharge value queue and performs pre-processing, defines each sub-space of the partial discharge value of each sampling time point after pre-processing, obtains the distortion amplitude of the partial discharge value of each sampling time point in each sub-space, obtains the distortion stability of the partial discharge value of each sampling time point in each sub-space according to the distortion amplitude, obtains the distortion correctness of the partial discharge value of each sampling time point according to the distortion stability and the arrangement of the partial discharge value in the sub-space, obtains the correct distortion amplitude of the partial discharge value of each sampling time point by combining the distortion correctness and the distortion amplitude, and achieves that if the collected partial discharge value is higher than the defined critical value, the alarm connected to the monitoring server will alarm, and the establishment method of the correct distortion amplitude of the partial discharge value of each sampling time point is:

[0049] If the individual partial discharge value is different from the overall trend, it means that the partial discharge value at the sampling time point is disturbed by a large amount of noise value, resulting in a low error. The greater the damage of the individual partial discharge value to the trend, the higher the distortion amplitude of the partial discharge value. The trend of the partial discharge value at different sampling time points is different, and the trend change of the partial discharge value depends on the sub-region where the partial discharge value is located. Therefore, the correctness of the distortion amplitude of the individual partial discharge value in different sub-regions is different.

[0050] The trend of the partial discharge value is more stable in the period of increasing capacity of the sub-region, and the correctness of the distortion amplitude of the partial discharge value is greater. The trend in the sub-region is more stable, and the distortion amplitude of the individual partial discharge value is more stable. Therefore, when the capacity of the sub-region changes, the distortion amplitude of the partial discharge value is more stable, and the correctness is greater.

[0051] The distortion noise value in the partial discharge value damages the trend of the sub-region, and the trend of the partial discharge value in the sub-region is often reflected by the regression performance of the partial discharge value in the sub-region.

[0052] For each partial discharge value after the pre-processing in the queue of the partial discharge values, a sub-region is formed by selecting a plurality of adjacent partial discharge values in the queue. Regression is performed on all the partial discharge values in the sub-region. The sampling time point is used as the X coordinate, and the partial discharge value is used as the Y coordinate. The regression line of all the partial discharge values in the sub-region is obtained by using the RANSAC method. Then, the median in the sub-region is removed, and the regression line of the remaining all partial discharge values is obtained by using the RANSAC method. The regression offset of the regression line is calculated, that is, the regression offset of the regression line of all the partial discharge values in the sub-region and the regression offset of the regression line of the remaining all partial discharge values. The regression offset is defined as regression offset one and regression offset two, respectively.

[0053] In a preferred but non-limiting embodiment of the present application, the regression offset is the amount obtained by subtracting the partial discharge value at the sampling time point on the regression line from the partial discharge value obtained by sampling at the sampling time point.

[0054] In the present application, the partial discharge value corresponding to each sampling time point after the pre-processing in the queue of the partial discharge values is sequentially taken as a left and right partial discharge value in the queue of the partial discharge values. The number of the partial discharge values can be twelve.

[0055] In a preferred but non-limiting embodiment of the present application, Step 2 further comprises: establishing the distortion amplitude of the partial discharge value at each sampling time point according to the regression offset, and the equation is: ​

[0056] Equation, represent the first the distortion amplitude of the local emission value at the sampling time point in the highest sub-space, represent the first the regression offset of the local emission value at the sampling time point in the highest sub-space, represent the first the regression offset of the local emission value at the sampling time point in the highest sub-space, represent the number of local emission values contained in the highest sub-space, represent the local emission values at the sampling time point, perform standardization.

[0057] remove the regression offset of the local emission value at the sampling time point from the highest sub-space, the regression offset of the local emission value at the sampling time point in the highest sub-space, the trend error of the local emission value at the sampling time point in the highest sub-space, the distortion amplitude of the local emission value at the sampling time point in the highest sub-space.

[0058] In a preferred but non-limiting embodiment of the present application, Step 2 further comprises: when the distortion of the local emission value is represented by the damage amplitude of the sub-space trend, the number of local emission values in the sub-space plays a key role in representing the trend change, and different sub-spaces with different numbers of local emission values have different trend presentations, so, to accurately represent the distortion of the local emission value, the present application defines different capacity sub-spaces for each local emission value at the sampling time point, and the number of local emission values in the sub-space is , where represents the query frequency, represents the query direction, in the present application, the value of is two, the value of is two, three, four, five, six, seven, eight, nine, ten, eleven, and for each local emission value at the sampling time point, ten different capacity sub-spaces are identified through ten queries;

[0059] The distortion amplitude of each local emission value at the sampling time point in each sub-space is obtained by using the same calculation method as the distortion amplitude of the local emission value in the highest sub-space.

[0060] For each sub-space of each local emission value at the sampling time point, when performing regression on the local emission value in the sub-space to represent the trend change of the sub-space, the change of the capacity of the sub-space determines the trend presentation of the sub-space, which in turn determines the distortion identification of the local emission value.

[0061] In a preferred but non-limiting embodiment of the present invention, Step 2 further includes: facing the first The partial discharge value at the sampling time point is used to establish the distortion stationarity within each spatial region, and its equation is: Within the equation, Representing the The partial discharge value at the sampling time point is at the first The distortion stability of each subspace, Representing the The partial discharge value at the sampling time point is at the first The distortion amplitude of each spatial component Representing the The partial discharge value at the sampling time point is at the first The distortion amplitude of each spatial component This represents the pre-defined running span, and its value is... .

[0062] As the capacity of the partial discharge space increases, the trend of the partial discharge value changes more smoothly in the partial discharge space, the corresponding regression offset is lower, the variation of the distortion amplitude of the partial discharge value becomes more stable, and the distortion amplitude of the partial discharge value in the corresponding partial discharge space is more accurate.

[0063] During periods of increased partial discharge space capacity, new clutter values ​​are often introduced, causing the partial discharge space trend to decline. The partial discharge value regression offset in the partial discharge space increases. Therefore, the greater the increase in the partial discharge space regression offset during periods of increased partial discharge space capacity, the higher the probability of introducing new clutter values. Consequently, the partial discharge value distortion amplitude reflected by the partial discharge space becomes more inaccurate. Simultaneously, the greater the stability of the partial discharge value distortion amplitude along with the change in partial discharge space capacity, the more accurate the distortion amplitude of the partial discharge value becomes.

[0064] In a preferred but non-limiting embodiment of the present invention, the distortion accuracy of the partial discharge value at each sampling time point in each spatial region is established based on the above analysis, and the equation is: Within the equation, Representing the The partial discharge value at the sampling time point is at the first The distortion accuracy of each subspace, Representing the The partial discharge value at the sampling time point is at the first The distortion stability of each subspace, Representing the The partial discharge value at the sampling time point The regression offset of all partial discharge values ​​within each subspace. Representing the The partial discharge value at the sampling time point Regression offset of the local emission value in each sub-space, representing the corrected distortion value of the local emission value at the sampling time point, representing the number of sub-spaces containing the local emission value, representing the number of sub-spaces containing the local emission value, representing the standardization by Z-score method,

[0065] In the preferred but non-limiting embodiment of the present application, the correct distortion value of the local emission value at each sampling time point in each sub-space is established by combining the distortion amplitude and the distortion correctness, and the equation is: representing the corrected distortion value of the local emission value at the sampling time point, representing the corrected distortion value of the local emission value at the sampling time point, representing the distortion correctness of the local emission value at the sampling time point in the sub-space, representing the distortion amplitude of the local emission value at the sampling time point in the sub-space, representing the distortion amplitude of the local emission value at the sampling time point in the sub-space, representing the number of sub-spaces of the local emission value at the sampling time point.

[0066] Step 3: Obtain the corrected distortion value of the local emission value at all sampling time points, and thus select the distortion value;

[0067] In the preferred but non-limiting embodiment of the present application, Step 3 specifically includes: obtaining the standardized value after the Z-score method is used to perform the standardization treatment on the corrected distortion value of the local emission value at all sampling time points, and then identifying the local emission value with a standardized value higher than a limited value as the distortion value. The value of the limited value can be four-fifths.

[0068] Step 4: Finally compare the local emission value after the noise removal with the defined critical value according to the local emission value after the distortion value is removed.

[0069] In the preferred but non-limiting embodiment of the present application, Step 4 specifically comprises: in the face of the entire partial discharge value after the previous treatment, removing the entire distortion value, then using the EM algorithm to perform the filling of the partial discharge value at the sampling time point after removing the partial discharge value to ensure the completeness of the partial discharge value, then taking the filled entire partial discharge value as the partial discharge value after removing the impurities, using the partial discharge value after removing the impurities compared with the defined threshold value, if the partial discharge value after removing the impurities is higher than the defined threshold value, the monitoring server alarms the alarm connected thereto.

[0070] As shown in Figure 2, the partial discharge value monitoring intelligent diagnosis device of the power distribution station room according to the present application comprises:

[0071] The partial discharge monitoring terminal arranged on the component of the power distribution station room is connected to the monitoring server through the communication equipment, the monitoring server is connected to the server, and the partial discharge monitoring terminal is used to transmit the partial discharge value of the component of the power distribution station room sampled to the monitoring server through the communication equipment. The monitoring server performs the removal of the impurities after collecting the partial discharge value, and then compares the partial discharge value after removing the impurities with the defined threshold value. If the partial discharge value after removing the impurities is higher than the defined threshold value, the monitoring server alarms the alarm connected thereto; the component of the power distribution station room can be the gas insulated high-voltage cabinet of the switch room of the power distribution station room. The power distribution station room is the power distribution station room, which includes the switching station, the power distribution room, the ring network room, the ring network box and the box-type substation.

[0072] The modules running on the monitoring server include:

[0073] The previous treatment module is used to collect the partial discharge value queue and perform the previous treatment;

[0074] The stability module is used to obtain the distortion amplitude of the partial discharge value at each sampling time point in each partial space after the previous treatment, obtain the distortion stability of the partial discharge value at each sampling time point in each partial space according to the distortion amplitude, obtain the distortion correctness of the partial discharge value at each sampling time point in each partial space according to the distortion stability, and obtain the correct distortion amplitude of the partial discharge value at each sampling time point;

[0075] The selection module is used to obtain the correct distortion amplitude of the partial discharge value at all sampling time points, thereby selecting the entire distortion value;

[0076] The removal module is used to finally compare the partial discharge value after removing the impurities with the defined threshold value according to the partial discharge value after removing the distortion value.

[0077] The present application has the following technical effects compared with the prior art:

[0078] The trend change of the partial space of each sampling point is analyzed, the distortion amplitude of the partial space of each sampling point is established, the difference between the overall trend change of the partial space and the partial space of each sampling point is related, the amplitude of the partial space of each sampling point disturbed by the clutter value is reflected, the distortion stability of the partial space of each sampling point is obtained according to the distortion amplitude, the stability of the distortion amplitude of each sampling point is reflected with the change of the capacity of the partial space, the trend change condition of several partial spaces is related, the distortion correctness of the partial space of each sampling point is obtained according to the distortion stability, the distortion stability of the partial space of each sampling point and the regression offset of the partial space of each sampling point are related, the safety of the distortion amplitude of the partial space is improved, the correct distortion amplitude of the partial space of each sampling point is obtained, the distortion value is removed by using the correct distortion amplitude, and the partial space after removing the clutter is compared with the defined critical value, if the partial space after removing the clutter is higher than the defined critical value, the alarm connected with the monitoring server alarms, the partial space has the effects of good correctness, good precision and good safety.

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, not to limit it. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent replacements can still be made to the specific embodiments of the present application without departing from the spirit and scope of the present application. Any modification or equivalent replacement should be covered in the protection scope of the claims of the present application.

Claims

1. A method for intelligent diagnosis of local value monitoring of a power station room, characterized in that, The method for removing the impulsive noise from the collected partial discharge values comprises the following steps: The partial discharge monitoring terminal transmits the sampled partial discharge values of the components in the substation room to the monitoring server through the communication device, the monitoring server removes the impulsive noise from the collected partial discharge values, and then compares the partial discharge values after removing the impulsive noise with the defined threshold value. If the partial discharge values after removing the impulsive noise are higher than the defined threshold value, the monitoring server triggers the alarm connected thereto to alarm. The method for removing the impulsive noise from the collected partial discharge values comprises the following steps: Step 1: Collecting the partial discharge value queue and performing the pre-processing; Step 2: Obtaining the distortion amplitude of the partial discharge value of each sampling time point in each sub-space, obtaining the distortion stability of the partial discharge value of each sampling time point in each sub-space according to the distortion amplitude, obtaining the distortion correctness of the partial discharge value of each sampling time point in each sub-space according to the distortion stability, and obtaining the correct distortion amplitude of the partial discharge value of each sampling time point; Step 3: Obtaining the correct distortion amplitude of the partial discharge value of all sampling time points, and thus selecting all the distortion values; Step 4: Comparing the partial discharge values after removing the impulsive noise with the defined threshold value according to the partial discharge values after removing the distortion values.

2. The method of claim 1, wherein, In Step 1, the partial discharge monitoring terminal samples the partial discharge values of the components in the substation room and transmits the partial discharge values and their sampling time points to the monitoring server in real time according to the sequence of the sampling time points. The monitoring server arranges the collected partial discharge values into a partial discharge value queue according to the sequence of the sampling time points.

3. The method of claim 2, wherein the method further comprises: In Step 1, the partial discharge values are pre-processed by using the moving average method, and the partial discharge values after the pre-processing are obtained.

4. The method of claim 3, wherein the method further comprises: Step 2 specifically comprises the following steps: For each partial discharge value after the pre-processing in the partial discharge value queue, a sub-space is selected by using a plurality of adjacent partial discharge values in the queue, then regression is performed on all the partial discharge values in the sub-space, the sampling time point is used as the X coordinate and the partial discharge value is used as the Y coordinate, the regression line of all the partial discharge values in the sub-space is obtained by using the RANSAC method, then the median of the sub-space is removed, the regression line of the residual partial discharge values is obtained by using the RANSAC method, and the regression offset of the regression line, i.e. the regression offset of the regression line of all the partial discharge values in the sub-space and the regression offset of the regression line of the residual partial discharge values, is defined as the regression offset one and the regression offset two, respectively; The regression offset is the quantity obtained by subtracting the partial discharge value of the sampling time point on the regression line from the partial discharge value sampled at the sampling time point; corresponding to the partial discharge value at each sampling time point after the preceding treatment in the partial discharge value queue, the partial discharge value queue is sequentially taken from left to right respectively The sub-space of the partial discharge value of each sampling time point, which is defined as the highest sub-space, is the partial discharge value of the sampling time point.

5. The method of claim 4, wherein the method further comprises: Step 2 further includes: establishing the distortion amplitude of the partial discharge value at each sampling time point based on the regression offset, the equation of which is: Within the equation, Representing the The distortion amplitude of the partial discharge value at the sampling point in the highest spatial distribution. Representing the The regression offset of the partial discharge value at the sampling time point in the highest distribution space is one. Representing the The regression offset of the partial value at the sampling point in the highest distribution space is two. This represents the number of partial discharge values ​​contained within the highest-level distribution space. The representative used the Z-score method to... Implement standardization.

6. The method of claim 5, wherein the method further comprises: Step2 specifically further comprises: facing each sampling point of partial discharge value, define different capacity of sub-space, sub-space contains the number of partial discharge value is , here represents the query frequency, represents the query direction, the value is two, the value is two, three, four, five, six, seven, eight, nine, ten, eleven, facing each sampling point of partial discharge value, identified ten different capacity of sub-space through ten queries; The distortion amplitude of the partial discharge value of each sampling time point in each sub-space is obtained by using the same calculation method as the distortion amplitude of the partial discharge value in the highest sub-space.

7. The method of claim 6, wherein the method further comprises: Step 2 specifically further comprises: facing the partial space of the first The partial space of the first The partial space of the first The partial space of the first The partial space of the first The partial space of the first The partial space of the first The partial space of the first The partial space of the first The partial space of the first The partial space of the first The partial space of the first The partial space of the first 8. The method of claim 6, wherein the method further comprises: The equation for establishing the distortion accuracy of the partial discharge value at each sampling time point in each partial space is: In the equation, represents the distortion accuracy of the partial discharge value at the sampling time point in the first partial space, represents the distortion accuracy of the partial discharge value at the sampling time point in the first partial space, represents the distortion accuracy of the partial discharge value at the sampling time point in the first partial space, represents the regression deviation of the partial discharge value at the sampling time point in the first partial space, represents the regression deviation of the partial discharge value at the sampling time point in the first partial space, represents the regression deviation of the partial discharge value at the sampling time point in the first partial space, represents the number of partial discharge values contained in the first partial space at the sampling time point, represents the number of partial discharge values contained in the first partial space at the sampling time point, represents the number of partial discharge values contained in the first partial space at the sampling time point, represents the standardization performed on the Z-score method; The partial discharge value at each sampling time point is established as the correct distortion amplitude of each sub-space by combining the distortion amplitude and distortion accuracy, and the equation is: within the equation, Representative of the first correctly distorted amplitude of the partial discharge value at the sampling point, Representative of the first The PD values at the sampling points were in the 1st the distortion accuracy of the individual subspaces, Representative of the first The PD values at the sampling points were in the 1st the amplitude of the distortion of the subfield space, Representative of the first The number of the sub-spaces of the partial discharge value of the sampling time point.

9. The method of claim 8, wherein, Step 3 specifically includes: after starting the correct distortion amplitude of partial discharge value of all sampling points is standardized by Z-score method, the standardized value is obtained, and then the partial discharge value with standardized value higher than the limited value is identified as the distortion value. ​ The step 4 specifically comprises: facing the whole partial discharge value after the pre-treatment, removing the whole distortion value, then using the EM algorithm to fill the partial discharge value at the sampling time point after removing the partial discharge value, then taking the filled whole partial discharge value as the partial discharge value after removing the impurities, using the partial discharge value after removing the impurities to compare with the defined threshold value, if the partial discharge value after removing the impurities is higher than the defined threshold value, the monitoring server makes the alarm connected thereto alarm.

10. A substation room local value monitoring intelligent diagnosis device, characterized in that, The method comprises: The partial discharge monitoring terminal arranged on the component in the power distribution station room is connected with the monitoring server through the communication equipment, the monitoring server is connected with the server, the partial discharge monitoring terminal is used to transmit the partial discharge value of the component in the power distribution station room sampled to the monitoring server through the communication equipment, the monitoring server removes the impurities from the received partial discharge value, then compares the partial discharge value after removing the impurities with the defined threshold value, if the partial discharge value after removing the impurities is higher than the defined threshold value, the monitoring server makes the alarm connected thereto alarm; The module running on the monitoring server comprises: The pre-treatment module is used to receive the partial discharge value queue and perform the pre-treatment; The stability module is used to obtain the distortion amplitude of the partial discharge value at each sampling time point in each partial space after the pre-treatment, obtain the distortion stability of the partial discharge value at each sampling time point in each partial space according to the distortion amplitude, obtain the distortion correctness of the partial discharge value at each sampling time point in each partial space according to the distortion stability, and obtain the correct distortion amplitude of the partial discharge value at each sampling time point; The selection module is used to obtain the correct distortion amplitude of the partial discharge value at all sampling time points, thereby selecting the whole distortion value; The removal module is used to finally compare the partial discharge value after removing the impurities with the defined threshold value according to the partial discharge value after removing the distortion value.

Images